-  
-   Notifications  You must be signed in to change notification settings 
- Fork 19.2k
DOC: update vendored numpydoc version #20383
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DOC: update vendored numpydoc version #20383
Conversation
| Hello @jorisvandenbossche! Thanks for updating the PR. 
 
 
 
 
 
 Comment last updated on March 27, 2018 at 14:57 Hours UTC | 
72e2106 to 67259df   Compare   | self.use_plots = config.get('use_plots', False) | ||
| self.use_blockquotes = config.get('use_blockquotes', False) | ||
| self.class_members_toctree = config.get('class_members_toctree', True) | ||
| self.attributes_as_param_list = config.get('attributes_as_param_list', True) | 
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Will this ensure that Attributes is formatted like a table?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes (updated the top post). For now, that is the only change I made to the numpydoc source (it's in a separate commit).
But, if I am the only one that finds that better, I am OK to leave that.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No, I prefer the attributes table. I just didn't realize there was an option for it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Well, there isn't ;-) That's something I added here (but with only a few lines: 075438f), and that's what I proposed to do in numpy/numpydoc#106, so it would be nice to have this in upstream)
| We still have the problem we fixed with this hack: 70c9d31, related to the property's that are None. However, due to recent work, I think only  | 
| 
 | 
| Codecov Report
 @@ Coverage Diff @@ ## master #20383 +/- ## ========================================== - Coverage 91.85% 91.82% -0.03%  ========================================== Files 152 152 Lines 49233 49233 ========================================== - Hits 45222 45210 -12  - Misses 4011 4023 +12
 
 Continue to review full report at Codecov. 
 | 
efab591 to faba146   Compare   | Updated this. The remaining related warnings I think are: /tmp/doc/source/generated/pandas.Index.holds_integer.rst: WARNING: document isn't included in any toctree this seems to be because they are not listed manually in api.rst, and are also not linked in the class docstring page (because they have no docstring they appear as plain table entry, not link to the docstring page. With as a result there is nowhere a link to that page). | 
| Would adding them to the  | 
| 
 So now I need to remember again why we had those entries there at the bottom of api.rst ... Which you added in #18202. Because this are attributes that were not yet included in api.rst before, and it is only now because they were no longer included in their class docstring page, there was a warning. | 
Numpydoc upstream at 21a194e9b42ebe95a70475e35cb69bcb4801ff36
…riptor to check attributes This is needed to ensure .columns / .index / cached properties are recognized as an attribute
faba146 to 41bbc7b   Compare   | OK, so it seems that as long as they are properly included in the class docstring page toctree, that is also fine, so in principle then we can get rid of the full list of hidden items in api.rst (but for that, we need to add docstrings to some of those methods). That's the proper fix for the warnings, but will leave that for another PR (now added them to the hidden list in api.rst) | 
| Travis doc build looks good | 
The first commit only updates our vendored version with the upstream one.
Additional commits make it work with our docs:
numpydoc_use_blockquotes=Falsefor now for compatibility (it is deprecated). To fix this, we need to fix: css styling of description lists + warnings generated due to**kwargsIt is probably worth to look at the separate commits (and would also like to keep them when merging)